We are trying to ascertain the molecular basis underlying actin filament formation and turnover. Actin must form a filament to fulfill its biological roles. It must bind to ATP or ADP along with a divalent cation to maintain its native structure. Finally, the hydrolysis of the nucleotide concomitant with polymerization and subsequent release of the Pi creates ADP-F actin which is a less stable filament than ATP- or ADP- Pi F actin. Thus, actin filament cycling may depend on the rate of Pi release from the filament. Although the structure of G-actin is known to atomic resolution, the filament structure has not been directly determined. Holmes et al. proposed a model of F-actin based on the structure of the monomer coupled with fiber diffraction data from oriented actin gels. The model predicts a reorientation of a loop between subdomains 3 and 4, containing a hydrophobic "plug", to allow it to interact with a hydrophobic pocket composed of subdomains 2 and 4 of adjacent monomers in the opposing strand. Such a bridge would greatly enhance interstrand stabilization of the structure. However, this aspect of the model has not been experimentally verified. We have altered the plug of yeast actin by site-directed mutagenesis to begin to test the plug-pocket interaction theory. To continue, we will place a cysteine in a proposed conformationally active plug position, L269. We will attach a fluorescent probe and study the behavior of the plug in the G and F states using steady state fluorescence and fluorescence resonance energy transfer. We will create mutants at H73 to test a theory proposing this residue as an important factor in determining the rate of Pi release. The cold-sensitive polymerization defect associated with these mutants suggests they may weaken the subdomain 2 component of the pocket of the plug-pocket interaction. We will test this by examining the factors needed to stabilize these mutant filaments against the cold. We will combine the H73 and L269C mutations to examine the behavior of the loop in the monomer under conditions where it would normally polymerize. Finally, we will use genetics to determine which proteins, in vivo, restore polymerizability to our polymerization- defective mutants. This work should provide new insight concerning the molecular basis underlying F-actin assembly and disassembly.